Structure and stability of Co(II)-complexes formed by wild-type and metal-ligand substitution mutants of T4 gene 32 protein

Abstract

Phage T4 gene 32 protein (gp32) is a zinc metalloprotein that binds cooperatively and preferentially to single-stranded nucleic acids and functions as a replication and recombination accessory protein. We have previously shown that the ZN(II) coordination by gp32 employs a metal ligand donor set unrelated to any known zinc-finger motif thus far described and is derived from the His64-XI2-Cys77-Xg-Cys87-X2-CYS90 sequence in the ssDNA-binding core domain of the molecule. Crystallographic studies reveal that His64 and Cys77 are derived from two independent p-strands and are relatively more buried from solvent than are Cys87 and Cys9O, which combine to nucleate an (X-helix. In an effort to understand the origin of the stability of the metal complex, we have employed an anaerobic optical spectroscopic, competitive metal binding assay to determine the coordination geometry and association constants (Ka) for the binding of CO(II) to wild-type gp32 and a series of zinc ligand substitution mutants. We find that all non-native metal complexes retain tetrahedral coordination geometry but are greatly destabilized in a manner essentially independent of whether a new protein-derived coordination bond is forfned (e.g., in H64C gp32) or not. Quantitative Co(H) binding isotherms for the His64 mutants reveal that these gp32s form a dimeric CYS4 tetrathiolate intermediate complex of differing affinities at limiting [Co]f; each then rearranges at high [Co]f to form a monomolecular site of the expected geometry and Ka=IXIO4 M-1. C87S and C90A gp32s, in contrast, form a single complex at all [Co]f, consistent with CYS2-His-H20 tetrahedral geometry of Ka=1-2xlo5 M-1. The susceptibility of all mutant metal sites to oxidation by 02 is far greater than the wild-type protein; none appear to be functional ssDNA binding proteins. These studies reveal that the local protein structure greatly limits accommodation of an altered complex in a ligand-specific manner. The implications of this work for de novo design of zinc complexes in proteins will be discussed.